Concept-based Entity Similarity Search

• Main Authors: Hsin-Chung Lin, 林信仲
• Other Authors: Pu-Jen Cheng
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/64760378663833646933

碩士 === 國立臺灣大學 === 資訊網路與多媒體研究所 === 96 === In this paper we address the problem of concept-based entity similarity search. In entity similarity search, given a query and an entity type, a search system returns a ranked list of entities in the type (e.g., person name, e-mail) relevant to the query. Ranking is a key issue in entity similarity search. In literature, entity extraction focuses on how to extract correct entities and entity ranking focuses on the ranking of entities according to the relevance between entities. In general, many features may be useful for ranking in entity similarity search no more than the contextual feature. We propose a general framework for entity similarity search on the web. And this framework is able to adjust the similarity function according to the user’s relevance feedback. The assumption of this problem, we propose there are semantic relationships among entities at conceptual level. We evaluate our online prototype over a Web corpus, and show that our approach performs effectively.